(A) The hypothesis that a blood vessel's structure is determined by the target tissue it supplies and not by the vessel's source, does not apply to mature tissue but rather, to fetal tissue. Instead of adult skeletal muscle autografts being exclusively vascularized by muscle type of continuous vessels (CV), grafts also contain fenestrated vessels (FV) of mature choroid plexus; FV are retained rather than being totally replaced by CV. The hypothesis does hold for fetal tissue up to a certain stage of development. E14 and E16 fetal muscles were grafted to the choroid plexus of adult hosts. In E14 grafts, about 80% of the muscle capillaries were CV, like those of brain or muscle. Very few of the CV were, however, from brain because there was no immunostaining of brain barrier antigen. In E16 grafts, about 70% of the vessels were FV. It is concluded that only target tissue need be fetal for the hypothesis to apply. Donor fetal tissue is being labeled with bromodeoxyuridine and 3H-thymidine to identify vessel sources. (B) The mechanism of opening the blood-brain barrier (BBB) with RMP-7, a bradykinin analog, to small molecules such as sucrose and inulin (5 kD), is being examined by light and electron microscopy. Biotinylated dextran (3 kD) passes the barrier only very infrequently and in small amount whereas exudates of the smaller and highly cationic La3+ are much more consistent and escape by way of interendothelial clefts rather than vesicular transcytosis.